Optimized machine learning based comparative analysis of predictive models for classification of kidney tumors

The kidney is an important organ that helps clean the blood by removing waste, extra fluids, and harmful substances. It also keeps the balance of minerals in the body and helps control blood pressure. But if the kidney gets sick, like from a tumor, it can cause big health problems. Finding kidney is...

Full description

Saved in:
Bibliographic Details
Published in:Scientific reports Vol. 15; no. 1; pp. 30358 - 13
Main Authors: Anand, Vatsala, Khajuria, Ajay, Pachauri, Rupendra Kumar, Gupta, Vinay
Format: Journal Article
Language:English
Published: London Nature Publishing Group UK 19.08.2025
Nature Publishing Group
Nature Portfolio
Subjects:
639/166
692/699
Accuracy
Algorithms
Automation
Blood pressure
Cancer
Classification
Comparative analysis
Datasets
Decision Trees
Deep learning
Diseases
Feature selection
Health
Health problems
Humanities and Social Sciences
Humans
Kidney diseases
Kidney Neoplasms - classification
Kidney Neoplasms - diagnosis
Kidney Neoplasms - pathology
Kidney stones
Kidneys
Learning algorithms
Literature reviews
Lung cancer
Machine Learning
Medical research
Mortality
multidisciplinary
Patients
Prediction models
Radiomics
Regression analysis
Research methodology
Science
Science (multidisciplinary)
Support Vector Machine
Support vector machines
Transfer learning
Tumors
Transfer learning
Health
Mortality
Classification
Cancer
Diseases
ISSN:2045-2322, 2045-2322
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The kidney is an important organ that helps clean the blood by removing waste, extra fluids, and harmful substances. It also keeps the balance of minerals in the body and helps control blood pressure. But if the kidney gets sick, like from a tumor, it can cause big health problems. Finding kidney issues early and knowing what kind of problem it has is very important for good treatment and better results for patients. In this study, different machine learning models were used to detect and classify kidney tumors. These models included Decision Tree, XGBoost Classifier, K-Nearest Neighbors (KNN), Random Forest, and Support Vector Machine (SVM). The dataset splitting is done in two ways 80:20 and 75:25 and the models worked best with the 80:20 split. Among them, the top three models—SVM, KNN, and XGBoost—were tested with different batch sizes, which are 16 and 32. SVM performed best when the batch size was 32. These models were also trained using two types of optimizers, called Adam and SGD. SVM did better when using the Adam method. SVM had the highest accuracy of 98. 5%, then came KNN with 90.4%. This method will help healthcare professionals in the early diagnosis of disease.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:2045-2322
2045-2322
DOI:10.1038/s41598-025-15414-w